Artificial Intelligence Frameworks Research Articles

Explainable AI Frameworks: Navigating the Present Challenges and Unveiling Innovative Applications

This study delves into the realm of Explainable Artificial Intelligence (XAI) frameworks, aiming to empower researchers and practitioners with a deeper understanding of these tools. We establish a comprehensive knowledge base by classifying and analyzing prominent XAI solutions based on key attributes like explanation type, model dependence, and use cases. This resource equips users to navigate the diverse XAI landscape and select the most suitable framework for their specific needs. Furthermore, the study proposes a novel framework called XAIE (eXplainable AI Evaluator) for informed decision-making in XAI adoption. This framework empowers users to assess different XAI options based on their application context objectively. This will lead to more responsible AI development by fostering transparency and trust. Finally, the research identifies the limitations and challenges associated with the existing XAI frameworks, paving the way for future advancements. By highlighting these areas, the study guides researchers and developers in enhancing the capabilities of Explainable AI.

Diversity, Equity, and Inclusion, and the Deployment of Artificial Intelligence Within the Department of Defense

Artificial Intelligence (AI) adoption has seen substantial growth across industries. This paper explores the escalating use of AI within the United States Department of Defense (DoD) and the implications that diversity, equity, and inclusion (DEI) have on Service members and Civilians across the Department. More specifically, this paper explores the DEI considerations within AI technologies on individual, team, and Department readiness. The DoD's AI usage spans various strategic and operational capabilities, however this paper explores two critical domains: healthcare and recruitment. In healthcare, AI offers the promise of early disease detection, enhanced diagnostic capabilities, and streamlined administrative processes. However, potential biases stemming from homogenous training data threaten the accuracy and reliability of these systems, jeopardizing Service member health and eroding trust in AI-assisted medical decision-making and potentially the DoD at large. In recruitment, while AI promises efficiency in identifying ideal candidates, its deployment can perpetuate biases, especially when the training data used is not representative of all demographics. Despite efforts to design "unbiased" systems by excluding demographic data, such strategies may inadvertently overlook the unique challenges faced by marginalized communities, further entrenching existing disparities. Both case studies underscore the importance of considering DEI in the development and deployment of AI systems. As the DoD continues to integrate AI into its operations, this paper’s recommendations stress the necessity of continuous DEI assessment to ensure that AI serves as an asset rather than a liability. The authors recommend the following: 1. Data diversity & review 2. Continuous monitoring and calibration 3. Stakeholder engagement 4. Adoption of DEI requirements within Ethical AI Frameworks 5. Further research

Gravitating towards Artificial Intelligence on Anti-Money Laundering A PRISMA Based Systematic Review

In Bangladesh, like many other countries, the implementation of Anti-Money Laundering (AML) measures plays a crucial role in the fight against financial offences and the maintenance of the financial system's integrity. The integration of Artificial Intelligence (AI) is one of the technologies for AML that has emerged as a promising approach to enhance the detection and prevention of financial crimes. The researcher aims to investigate the current state of AML, the impact of AI, the roles involved, and the legislation to achieve economic stability for the transformation of Bangladesh into a smart nation by 2041. In this systematic review study, the PRISMA framework is utilized to provide a comprehensive analysis of existing literature on the application of AI in AML, with a focus on risks, laws, and implications, particularly within the context of Bangladesh. The review synthesizes findings from diverse sources (Web of Sciences, Scopus, PubMed, and DOAJ indexing articles and reports) including academic research, industry reports, and regulatory documents, to assess the effectiveness, challenges, and ethical considerations associated with AI-driven AML solutions. Through a structured search strategy and rigorous selection criteria, the researchers discovered a total of 299 research papers, consisting of 181 latest studies and 117 previous relevant studies identified and analyzed to elucidate key trends, methodologies, and gaps in the current landscape of AI-enabled AML practices. The review also offers insights and recommendations for policymakers, practitioners, and researchers interested in leveraging AI technologies to combat money laundering effectively while navigating regulatory frameworks and safeguarding privacy rights. By critically examining the intersection of AI, AML, and legal frameworks, this study contributes to the ongoing discourse on the role of technology in addressing financial crime challenges in Bangladesh and beyond.

Discussing the Role of Explainable AI and Evaluation Frameworks for Safe and Effective Integration of Large Language Models in Healthcare

Discussing the Role of Explainable AI and Evaluation Frameworks for Safe and Effective Integration of Large Language Models in Healthcare

223 An Artificial Intelligence Framework for Brain Tumor Delineation in Intraoperative Ultrasound

INTRODUCTION: Automated delineation of gliomas in intraoperative ultrasound (iUS) could provide timely information to surgeons. While artificial intelligence (AI) frameworks have shown promising results for delineating gliomas in preoperative MRI, automated delineation in iUS remains a challenging and open problem. METHODS: A deep learning algorithm was developed to automatically delineate gliomas in intraoperative ultrasound. Instead of using time-consuming manual delineations from experts, the algorithm was developed using MR-based automated delineations. First, a MR nnUnet was trained for delineating gliomas in 3D preoperative MRI using an annotated public dataset (N=1000; RSNA-ASNR-MICCAI-BraTS 2021). Then, this algorithm delineated gliomas in preoperative MRI of patients (N = 84) in whom pre-dural opening 3D iUS reconstructed from a tracked handheld 2D probe was also acquired. To obtain the MR-based delineations in iUS, preoperative MRI and iUS were aligned using NiftyReg. Finally, a 3D iUS nnUnet was trained using a random subset (N = 78) of this pseudo-annotated iUS dataset. The method was assessed on the remaining test set (N = 943 slices from 6 volumes), with a treating neurosurgeon delineating gliomas in 2D iUS only and then in 2D iUS and aligned MR slices to obtain gold-standard delineations. RESULTS: The median Dice scores of the nnUnet and the neurosurgeon were respectively 78.6% (IQ:15.5%) and 84.3% (IQ: 21.1%) against the gold-standard delineations. In this preliminary study, the network's performance was comparable to the neurosurgeon's (two paired one-sided tests, equivalence margin:5%, p<1e-5, N = 943), with rapid delineation (<10 second) of 3D iUS (95-216 slices) compared to neurogeons (32-93 minutes), making automated delineation practical during surgery. CONCLUSIONS: The results demonstrate the feasibility of using AI-based methods for automated delineation of brain tumors in iUS, which could improve surgical outcomes by providing real-time information to surgeons.

Navigating and reviewing ethical dilemmas in AI development: Strategies for transparency, fairness, and accountability

As artificial intelligence (AI) continues to permeate various aspects of our lives, the ethical challenges associated with its development become increasingly apparent. This paper navigates and reviews the ethical dilemmas in AI development, focusing on strategies to promote transparency, fairness, and accountability. The rapid growth of AI technology has given rise to concerns related to bias, lack of transparency, and the need for clear accountability mechanisms. In this exploration, we delve into the intricate ethical landscape of AI, examining issues such as bias and fairness, lack of transparency, and the challenges associated with accountability. To address these concerns, we propose strategies for transparency, including the implementation of Explainable AI (XAI), advocating for open data sharing, and embracing ethical AI frameworks. Furthermore, we explore strategies to promote fairness in AI algorithms, emphasizing the importance of fairness metrics, diverse training data, and continuous monitoring for iterative improvement. Additionally, the paper delves into strategies to ensure accountability in AI development, considering regulatory measures, ethical AI governance, and the incorporation of human-in-the-loop approaches. To provide practical insights, case studies and real-world examples are analyzed to distill lessons learned and best practices. The paper concludes with a comprehensive overview of the proposed strategies, emphasizing the importance of balancing innovation with ethical responsibility in the evolving landscape of AI development. This work contributes to the ongoing discourse on AI ethics, offering a roadmap for navigating the challenges and fostering responsible AI development practices.

Theoretical frameworks in AI for credit risk assessment: Towards banking efficiency and accuracy

This paper delves into theoretical frameworks in AI for credit risk assessment, exploring how these frameworks enhance banking efficiency and accuracy. It discusses various AI techniques such as machine learning algorithms, neural networks, and natural language processing, and their application in credit risk assessment. Furthermore, it examines the challenges and opportunities presented by these frameworks, highlighting their potential to revolutionize the banking sector. Revolutionizing Credit Risk Assessment in Banking, The Role of Artificial Intelligence In the dynamic realm of finance, the assessment of credit risk stands as a fundamental pillar for banking institutions. Traditionally, this process has heavily relied on statistical models and historical data. However, the emergence of Artificial Intelligence (AI) has catalyzed a transformative shift in this domain. This paper elucidates the theoretical underpinnings of AI frameworks employed in credit risk assessment and investigates their profound implications for enhancing the efficiency and accuracy of banking operations. The exploration begins by delineating various theoretical frameworks in AI pertinent to credit risk assessment. Leveraging machine learning algorithms, neural networks, and natural language processing techniques, these frameworks offer innovative approaches to evaluate creditworthiness. Unlike conventional methods, AI-driven models possess the capacity to ingest vast datasets, identify intricate patterns, and adapt dynamically to evolving market dynamics. Such capabilities empower banks to make more informed and timely decisions regarding lending activities. Moreover, this paper delves into the practical application of AI techniques in credit risk assessment. Through case studies and empirical evidence, it elucidates how these advanced methodologies enable banks to mitigate risks while maximizing profitability. By harnessing AI, financial institutions can optimize credit scoring processes, identify potential defaulters with greater accuracy, and customize lending terms based on individual risk profiles. Additionally, AI facilitates real-time monitoring of credit portfolios, allowing proactive risk management and timely interventions to prevent adverse outcomes.

Analyzing Cryptocurrency Prices through Artificial Intelligence

Cryptocurrency is reshaping the financial landscape, gaining popularity and acceptance among merchants. Despite the increasing investments in cryptocurrency, its dynamic features, uncertainties, and predictability remain largely unknown, posing significant investment risks. This study aims to understand the factors influencing cryptocurrency value formation. Leveraging advanced artificial intelligence frameworks like fully connected Artificial Neural Network (ANN) and Long Short-Term Memory (LSTM) Recurrent Neural Network, we analyze the price dynamics of Bitcoin, Ethereum, and Ripple. Our findings indicate that ANN relies more on long-term history, while LSTM focuses on short-term dynamics, suggesting LSTM's superior efficiency in utilizing historical information. However, with sufficient historical data, ANN can achieve comparable accuracy to LSTM. This study underscores the predictability of cryptocurrency market prices, though the explanation may vary depending on the machine-learning model employed.

AI for smart cities opportunities and promising directions

The transformative potential of Artificial Intelligence (AI) in the realm of smart cities is an evolving landscape of innovation and challenges. This research undertook a comprehensive exploration of AI's impact, harnessing both quantitative and qualitative methodologies. Detailed analyses were performed on data from select smart cities globally, focusing on sectors such as energy, traffic, health services, and waste management. Additionally, perceptions and experiences of urban stakeholders were captured through interviews. The results solidified AI's tangible benefits in enhancing urban life quality, while also bringing forth concerns about data privacy, algorithmic biases, and socio-economic implications. The study concludes with a call for holistic AI frameworks, heightened public engagement, and interdisciplinary collaborations to ensure the ethical and sustainable evolution of AI-integrated smart cities.

An Agent-Based Model to Reproduce the Boolean Logic Behaviour of Neuronal Self-Organised Communities through Pulse Delay Modulation and Generation of Logic Gates.

The human brain is arguably the most complex "machine" to ever exist. Its detailed functioning is yet to be fully understood, let alone modelled. Neurological processes have logical signal-processing and biophysical aspects, and both affect the brain's structure, functioning and adaptation. Mathematical approaches based on both information and graph theory have been extensively used in an attempt to approximate its biological functioning, along with Artificial Intelligence frameworks inspired by its logical functioning. In this article, an approach to model some aspects of the brain learning and signal processing is presented, mimicking the metastability and backpropagation found in the real brain while also accounting for neuroplasticity. Several simulations are carried out with this model to demonstrate how dynamic neuroplasticity, neural inhibition and neuron migration can reshape the brain's logical connectivity to synchronise signal processing and obtain certain target latencies. This work showcases the importance of dynamic logical and biophysical remodelling in brain plasticity. Combining mathematical (agents, graph theory, topology and backpropagation) and biomedical ingredients (metastability, neuroplasticity and migration), these preliminary results prove complex brain phenomena can be reproduced-under pertinent simplifications-via affordable computations, which can be construed as a starting point for more ambitiously accurate simulations.

An integrated AI framework for managing organizational risk and climate change concerns in B2B market

The United Nations identified sustainable development goals that focus on the three pillars of sustainability-Economy, Society, and Environment. The capability of Artificial intelligence (AI) for societal good and support to these three pillars is a matter of research importance. Drawing attention to literature on various AI frameworks for climate change and the dynamic capability theory, this study conducts a survey of managers working within their AI ecosystem of B2B firms. Confirmatory factor analysis and structural equation modeling are used to test the reliability, validity, and fitness of the research model. This model illustrates how big data-based AI system can influence business performance due to climate change. Research findings reveals paths involving “big data based artificial intelligence system”, “climate change competency process (prediction competency, mitigation competency, adaptation competency)”, “climate competent decision-making,” and “firm performance”, are statistically significant. As a part of research validation, three business cases are studied that focus on the decision-making capability of business under an uncertain environment. This study justifies that climate competent decision-making is a critical skill for businesses operating under uncertain environmental changes. This research brings out management perceptions generated due to climate change and discusses how they influence managers' behaviors that alter relationship dynamics and structure.

AIfES: A Next-Generation Edge AI Framework.

Edge Artificial Intelligence (AI) relies on the integration of Machine Learning (ML) into even the smallest embedded devices, thus enabling local intelligence in real-world applications, e.g. for image or speech processing. Traditional Edge AI frameworks lack important aspects required to keep up with recent and upcoming ML innovations. These aspects include low flexibility concerning the target hardware and limited support for custom hardware accelerator integration. Artificial Intelligence for Embedded Systems Framework (AIfES) has the goal to overcome these challenges faced by traditional edge AI frameworks. In this paper, we give a detailed overview of the architecture of AIfES and the applied design principles. Finally, we compare AIfES with TensorFlow Lite for Microcontrollers (TFLM) on an ARM Cortex-M4-based System-on-Chip (SoC) using fully connected neural networks (FCNNs) and convolutional neural networks (CNNs). AIfES outperforms TFLM in both execution time and memory consumption for the FCNNs. Additionally, using AIfES reduces memory consumption by up to 54% when using CNNs. Furthermore, we show the performance of AIfES during the training of FCNN as well as CNN and demonstrate the feasibility of training a CNN on a resource-constrained device with a memory usage of slightly more than 100 kB of RAM.

Synthesizing Radiological Insights: Enhancing Lung Disease Classification through Multimodal Imaging

Precisely categorizing lung diseases is essential for effective medical treatments. This paper presents a comprehensive analysis of advanced methods in lung disease classification, with a focus on integrating diverse imaging techniques like computerized tomography (CT), X-rays, and magnetic resonance imaging (MRI). These imaging approaches collectively enhance the understanding of pulmonary conditions, aiding in early detection and differential diagnosis. The paper initially explains the fundamental principles of CT, MRI, and X-rays, highlighting their unique characteristics and roles in elucidating lung structures. It explores state-of-the-art methodologies, encompassing both traditional machine learning using engineered features and the expanding domain of deep learning utilizing neural networks to classify intricate diseases. A wide range of prevalent lung ailments, spanning from pneumonia and lung cancer to chronic obstructive pulmonary disease (COPD), are covered. Each domain delves into the considerations for adapting imaging modalities, involving data pre-processing, feature extraction, and algorithmic orchestration. Comparative evaluations of performance metrics offer insights into the effectiveness and limitations of each approach. Furthermore, the paper outlines the challenges associated with classifying lung diseases, including limited annotated data, complexities in model interpretation, and the seamless integration of algorithmic outcomes into clinical practices. As for future research avenues, the paper suggests innovative directions such as data augmentation, integrating multi-modal imaging information, and advancing transparent artificial intelligence (AI) frameworks to enhance their acceptance in clinical settings.

Synergizing Solar Cell Technology, Radio Waves, AI, and Their Impact on Business and Medicine

This analysis examines how solar cell technology, radio waves, and artificial intelligence (AI) are transforming a variety of industries, including business and medicine. Innovation, efficiency, and sustainability are all promised by the combination of these technologies. The trip starts with a look at each technology's historical foundations and development, highlighting its contributions to the revolutions in energy production, healthcare, and communication. Radio waves have changed over time in the context of wireless communication, from early broadcasting to 5G networks, which have ushered in a new era of connectedness. The interplay between radio waves and solar cell technology is especially noticeable in energy harvesting and wireless communication, making sustainable and effective power sources for off-grid places and Internet of Things devices possible. AI's introduction enhances these technologies even more, improving signal processing and spectrum management and bringing predictive maintenance to wireless networks. The application of AI raises ethical questions, especially in the areas of data management, privacy, and security. Sustainable solutions are required due to the influence these technologies have on the environment. An alternative that uses renewable energy sources, lowers operating costs, and leaves fewer carbon footprints is the use of solar-powered AI systems. These systems are used in data centers, manufacturing, smart buildings, and agriculture, promoting sustainability and efficiency in commercial activities. The combination of solar power, artificial intelligence, and radio waves is altering drug development, improving patient care, and revolutionizing medical diagnostics in the healthcare industry. AI-assisted medical diagnostics has the ability to detect diseases early and accurately, especially in the fields of pathology and medical imaging. Telemedicine is bolstered by wearables, solar-powered medical devices, and remote diagnostics, which allow for access to healthcare even in underserved and rural places, as well as remote patient monitoring. Interoperability, energy storage options, scalability, security, and data privacy are among the ongoing challenges. The ethical application of AI in medical diagnostics is crucial, as are environmental considerations in relation to the production and disposal of equipment. Research and development, standardization, cooperation, collaboration, teaching, sustainability projects, and the creation of ethical AI frameworks are some of the future directions.

Bipolar Fuzzy Multi-Criteria Decision-Making Technique Based on Probability Aggregation Operators for Selection of Optimal Artificial Intelligence Framework

Artificial intelligence (AI) frameworks are essential for development since they offer pre-built tools and libraries that speed up and simplify the production of AI models, leveraging symmetry to save time and effort. They guarantee effective computing by modifying code for particular hardware, facilitating quicker testing and deployment. The identification of a suitable and optimal AI framework for development is a multi-criteria decision-making (MCDM) dilemma, where the considered AI frameworks for development are evaluated by considering various criteria and these criteria may have dual aspects (positive and negative). Thus, in this manuscript, we diagnosed a technique of MCDM within the bipolar fuzzy set (BFS) for identification and selection of optimal AI framework for development. In this regard, we diagnosed probability aggregation operators (AOs) within BFS, such as probability bipolar fuzzy weighted averaging (P-BFWA), probability bipolar fuzzy ordered weighted averaging (P-BFOWA), immediate probability bipolar fuzzy ordered weighted averaging (IP-BFOWA), probability bipolar fuzzy weighted geometric (P-BFWG), probability bipolar fuzzy ordered weighted geometric (P-BFOWH), and immediate probability bipolar fuzzy ordered weighted geometric (IP-BFOWG) operators. The diagnosed technique would be based on these invented probably AOs. Afterward, in this manuscript, we took a case study and obtained the optimal AI framework for development by employing the diagnosed technique of MCDM. We also investigated the comparison of the devised theory with certain prevailing theories to reveal the dominance and significance of the devised theory.

From cancer big data to treatment: Artificial intelligence in cancer research.

In recent years, developing the idea of "cancer big data" has emerged as a result of the significant expansion of various fields such as clinical research, genomics, proteomics and public health records. Advances in omics technologies are making a significant contribution to cancer big data in biomedicine and disease diagnosis. The increasingly availability of extensive cancer big data has set the stage for the development of multimodal artificial intelligence (AI) frameworks. These frameworks aim to analyze high-dimensional multi-omics data, extracting meaningful information that is challenging to obtain manually. Although interpretability and data quality remain critical challenges, these methods hold great promise for advancing our understanding of cancer biology and improving patient care and clinical outcomes. Here, we provide an overview of cancer big data and explore the applications of both traditional machine learning and deep learning approaches in cancer genomic and proteomic studies. We briefly discuss the challenges and potential of AI techniques in the integrated analysis of omics data, as well as the future direction of personalized treatment options in cancer.

Abstract 18718: A Deep Learning Model for Classifying Left Ventricular Enlargement Using Transthoracic and Point-of-Care Echocardiograms

Background: Enlargement or dilation of the left ventricle (LV) is associated with increased cardiovascular morbidity and mortality, and is an established precursor of ventricular dysfunction and heart failure in almost all major cardiovascular conditions. Although there are existing artificial intelligence frameworks that make use of deep learning to detect LV enlargement, many entail LV segmentation, which can be error-prone especially for point-of-care ultrasound (POCUS) images. Goal: To develop a fully end-to-end deep learning framework that takes as input either transthoracic echocardiographic (TTE) or POCUS 2D image clips, and classifies directly (i.e., without first segmenting the LV) whether a patient has an enlarged LV. Methods: Using a sample of retrospective TTE data enriched for LV enlargement (out of 13,868 patients total, 3,351 had an enlarged LV defined using sex-specific guideline-directed criteria), we trained a classifier that takes as input apical 2-chamber and parasternal long axis videos, and provides as output a binary label denoting whether a patient has an enlarged LV. Then, we evaluated the trained model using a holdout testing dataset of retrospective TTE data (2,087 patients) as well as data from a prospective cohort of 428 patients from whom TTE and POCUS were collected simultaneously. Results: For the retrospective TTE testing dataset, the AUC was 0.851 (95% confidence interval (CI) of 0.830-871). For the prospective cohort, the model performed comparably for both TTE and POCUS video clips (TTE: AUC of 0.865, 95% CI of 0.823-0.906; POCUS: AUC of 0.840, 95% CI of 0.797-0.883). Conclusions: The current deep learning model performed comparably well for both TTE and POCUS echocardiograms. We regard these results as a proof-of-concept of the potential utility of this model in helping point-of-care clinicians make decisions regarding whether patients should receive further evaluation for LV enlargement and associated conditions.

Survey on Cloud Computing Integrated with Artificial Intelligence

The combination of cloud computing with Artificial Intelligence (AI) has become a viable approach in recent years to handle the increasing processing needs of AI applications. This article offers a thorough analysis of distributed and scalable artificial intelligence frameworks that use cloud computing to improve the effectiveness and performance of Deep Learning (DL). First, we provide a summary of common cloud services and AI frameworks, emphasizing the advantages and disadvantages of each. The important topics of data management and storage in cloud-based AI systems are then covered in detail, including feature engineering, data pretreatment, privacy, and security. Next, we investigate the distributed and parallel training methods for AI models, with particular emphasis on cloud-based training topologies, communication mechanisms, and model splitting. On the next chapters, we examine auto-scaling, load balancing, resource allocation, and performance benchmarking as optimization methodologies for AI workloads on the cloud. We also cover serverless deployment alternatives, containerization, best practices for monitoring, and the deployment and servicing of AI models on the cloud. We give a comprehensive study of costs, optimization methodologies, and case studies exhibiting successful installations to guarantee the cost-effectiveness of cloud-based AI systems. In conclusion, we provide an overview of the study's main conclusions, go through the difficulties and restrictions associated with cloud-based artificial intelligence, and point out new directions for future research in this area. For academics and practitioners trying to use cloud computing to create scalable, effective, and affordable artificial intelligence systems, this article will be a great resource.

Generative AI in Supply Chain Management

A new age of creativity and efficiency is ushered in by the integration of Generative Artificial Intelligence (AI) into supply chain management. This in-depth study examines the diverse effects of generative artificial intelligence on supply chain operations, including risk management, inventory optimization, procurement, logistics, and more. Given the predictive capacity of generative AI, traditional methods have been completely modified, enabling companies to anticipate demand, maximize inventory, and expedite procurement procedures with previously unheard-of accuracy. Real-time adaptation is made possible by its dynamic decision-making skills, which also help to promote resilience against interruptions and enable proactive reactions to changing market conditions. However, there are some challenges in implementing generative AI in supply chains. Obstacles requiring strategic navigation and organizational preparedness include skill gaps, ethical considerations, scalability issues, and data integration complexity. Future directions for generative artificial intelligence in supply networks are extremely promising. Substantial improvements are expected to be driven by advances in explainable AI, predictive analytics, seamless integration, and ethical frameworks. Redefining supply chain models could be facilitated by autonomous supply chains, adaptive resilience to disturbances, and increased transparency in decision-making.

Few-Shot Pixel-Precise Document Layout Segmentation via Dynamic Instance Generation and Local Thresholding.

Over the years, the humanities community has increasingly requested the creation of artificial intelligence frameworks to help the study of cultural heritage. Document Layout segmentation, which aims at identifying the different structural components of a document page, is a particularly interesting task connected to this trend, specifically when it comes to handwritten texts. While there are many effective approaches to this problem, they all rely on large amounts of data for the training of the underlying models, which is rarely possible in a real-world scenario, as the process of producing the ground truth segmentation task with the required precision to the pixel level is a very time-consuming task and often requires a certain degree of domain knowledge regarding the documents at hand. For this reason, in this paper, we propose an effective few-shot learning framework for document layout segmentation relying on two novel components, namely a dynamic instance generation and a segmentation refinement module. This approach is able of achieving performances comparable to the current state of the art on the popular Diva-HisDB dataset, while relying on just a fraction of the available data.